Your weekly Cutter Business Technology Journal Advisor

This Advisor is a deliverable of your subscription. View this message in a browser. 

Cutter Business Technology Journal

AI, ML, and Big Data: Functional Groups That Catch the Investor's Eye

by William Jolitz and Lynne Greer Jolitz
6 May 2020

Artificial intelligence (AI), machine learning (ML), and big data are expected to have a huge impact on how we live and what we choose to do. Most categories involving these technologies focus on specific lifestyle items like shopping, dining, movies, and so forth. But such customer-centric categories lack specificity from the investor’s eye. In this Advisor, we redraw the categories on which AI/ML and big data startups should focus, beginning with the first two categories that focus on functional structures:

  1. Surveillance — who you are, what you’re doing, where you are, what you want — requires heavy use of ML/AI to sort the wheat from the chaff for monetization. In 2020, we will see increasing use of AI/ ML in tracking (e.g., browser cookies, GPS, traffic location data), identity resolution (e.g., facial recognition, biometrics), and prediction (e.g., activities based on location, past affiliations interests, and recent activity). Models are judged by how many of these elements cohere to a person­alized lifestyle lensed through the aperture of data analysis.
  2. Entertainment is all about stickiness — where the customer’s eyeballs focus on one thing — whether watching a streaming movie or gaming or news or gossip. The low-latency infrastructure required for instantaneous AI/ML processing is intended to predict the customer interest intersections that are crucial to advertising and the creation of new content.

These two functional groups are both complementary and antagonistic in terms of use: com­plementary because a vendor or company can use surveillance to better target ads for entertainment, and antagonistic because entertainment is more a consumer of AI/ML than a creator of AI/ML. Entertainment must therefore anticipate interest to gain eyeballs because creation and production costs for content are high. In contrast, surveillance platforms, once they get going, are extremely cheap to run for the information ob­tained. The high investment cost is in gaining customers, but once accomplished, these platforms rely on inexpensive tweaks and gimmicks to retain those eyeballs as the platforms run on user-generated content. Surveillance platforms like social networks, search engines, or driving apps try to actively shape the user’s interest in real time, while entertainment has front-loaded the investment in the production of games, movies, and so on, and needs eyeballs to recoup that investment.

Continuing trends in ML include: (1) moving away from black box solutions, which fail to connect results to the means, via existing subject matter experts (the aim is traceability and transparency); (2) using compu­tationally densified processing (e.g., Google’s Transformer/Reformer) on deep search/deep neural networks (DNNs) and convolutional neural networks (CNNs)/recurrent neural networks (RNNs)/others that exploit the efficiencies of modern processor architectures; and (3) implementing tight go-to-market strategies that focus on taking out a single, high-profile, human resources–dominated business (or business function) in toto. Overall, trends in AI center around productivity improvement so as to do more with fewer human staff, while not overreaching to entirely eliminate, say, entry-level jobs.

Next comes the third functional category:

  1. Whitespace opportunities — wrestling with the complicated issues of climate change, energy, and transportation. These are fundamental areas that will require heavy use of ML, AI, and big data for data analysis, regulatory focus, and efficacy. The precedent in the previous decade is in aerospace and transportation with the success of SpaceX and and the more well-known Tesla, two companies founded by Elon Musk that have completely reshaped those industries. The rise of “super billionaires” in the past two decades of wealth creation has led to a greater willingness to buck traditional boundaries, whether held by old-style private industry blocks or by government fiat. 

[For more from the authors on this topic, see “Moving Forward in 2020: Technology Investment in ML, AI, and Big Data.”]

Recent content from Cutter Business Technology Journal contributors.

This Advisor is a service of your Cutter Consortium subscription. You can select or unselect which Advisors you receive from the Subscription tab of your account page at any time. 


Copyright © 2020 Cutter, an Arthur D. Little community | 10 High Street, Suite 900, Boston, MA 02110, USA | Manage your communication preferences.